RHIXO

RHIXO AI and Automation Terms

Terms for AI-assisted content generation, automation and Digital Associate functions.

This document forms part of the RHIXO controlled legal framework and applies as described below. It operates together with the RHIXO Customer Agreement, RHIXO Product and Service Schedule, accepted order and other incorporated documents where relevant.

1. Nature of AI outputs

AI systems are probabilistic and may produce inaccurate, incomplete, outdated, non-unique or unsuitable outputs. Output is not a representation of fact merely because it is generated confidently. Customers must apply appropriate human judgement and verification for their use case.

2. No professional advice

Unless a Product Schedule expressly states otherwise, AI output is not legal, tax, accounting, medical, engineering, safety, financial, employment or regulatory advice. The Customer remains responsible for obtaining professional advice when needed.

3. Customer inputs and authority

The Customer is responsible for inputs, instructions, uploaded material and context supplied to AI features and must have the rights and lawful basis needed to process them. Customers must not submit credentials, highly sensitive information or regulated data unless the feature is expressly designed and approved for that category.

4. Output rights and limitations

As between the parties and to the extent permitted by law, the Customer may use output generated specifically for it, subject to third-party rights, open-source licences, underlying platform rights and any Product Schedule restrictions. We do not warrant that output is unique, protectable, registrable or non-infringing.

5. Human oversight and irreversible actions

AI recommendations do not transfer decision responsibility from the Customer. Where a Service enables autonomous actions, permissions must be scoped to the minimum reasonably necessary. High-impact irreversible actions - including deletion of material data, external publication, material financial transactions, termination of user access, or other actions identified by the Product - must use an appropriate approval/confirmation control unless the Customer has expressly configured a governed alternative.

6. Digital Associates

Digital Associates may communicate, triage, draft, analyse, route, monitor and perform authorised workflows within configured authority. They are software-based workforce capabilities, not natural persons. Their persona or presentation does not override factual accuracy, governance, permissions or the Customer's responsibility for decisions.

7. Third-party AI providers

Some features may invoke third-party models or tools. Relevant data may need to be transmitted to those providers to perform the feature. Availability, model behaviour, pricing and terms may change. We will manage integrations we provide using commercially reasonable controls, but cannot guarantee third-party model output or continuous availability.

8. Model improvement and training

Customer Data will not be used to train a public foundation model unless the applicable terms clearly permit it and required authority or consent has been obtained. We may use de-identified operational telemetry, safety signals and feedback to operate, secure and improve our products where lawful.

9. Prohibited AI use

AI features must not be deliberately used for unlawful discrimination, fraud, impersonation, malicious code, rights infringement, unlawful surveillance, evasion of safety controls or other conduct prohibited by the AUP.

10. Safety controls and monitoring

We may apply automated and manual safeguards to detect abuse, security threats and policy violations. A safety control does not guarantee that every harmful or inaccurate output will be detected.

11. Changes and feature availability

AI models and capabilities may be upgraded, replaced, restricted or discontinued. Where a paid material capability is affected during a committed term, the Customer Agreement and Product Schedule govern notice and remedies.